Systematic review of shared decision-making interventions for people living with chronic respiratory diseases

Objective Shared decision-making (SDM) supports patients to make informed and value-based decisions about their care. We are developing an intervention to enable healthcare professionals to support patients’ pulmonary rehabilitation (PR) decision-making. To identify intervention components we needed to evaluate others carried out in chronic respiratory diseases (CRDs). We aimed to evaluate the impact of SDM interventions on patient decision-making (primary outcome) and downstream health-related outcomes (secondary outcome). Design We conducted a systematic review using the risk of bias (Cochrane ROB2, ROBINS-I) and certainty of evidence (Grading of Recommendations Assessment, Development and Evaluation) tools. Data sources MEDLINE, EMBASE, PSYCHINFO, CINAHL, PEDRO, Cochrane Central Register of Controlled Trials, the International Clinical Trials Registry Platform Search Portal, ClinicalTrials.gov, PROSPERO, ISRCTN were search through to 11th April 2023. Eligibility criteria Trials evaluating SDM interventions in patients living with CRD using quantitative or mixed methods were included. Data extraction and synthesis Two independent reviewers extracted data, assessed risk of bias and certainty of evidence. A narrative synthesis, with reference to The Making Informed Decisions Individually and Together (MIND-IT) model, was undertaken. Results Eight studies (n=1596 (of 17 466 citations identified)) fulfilled the inclusion criteria. Five studies included components targeting the patient, healthcare professionals and consultation process (demonstrating adherence to the MIND-IT model). All studies reported their interventions improved patient decision-making and health-related outcomes. No outcome was reported consistently across studies. Four studies had high risk of bias, three had low quality of evidence. Intervention fidelity was reported in two studies. Conclusions These findings suggest developing an SDM intervention including a patient decision aid, healthcare professional training, and a consultation prompt could support patient PR decisions, and health-related outcomes. Using a complex intervention development and evaluation research framework will likely lead to more robust research, and a greater understanding of service needs when integrating the intervention within practice. PROSPERO registration number CRD42020169897.


Introduction
-Could you clarify whether the results of recent systematic reviews on patient education materials have been evaluated to identify possibly relevant studies for inclusion? (E.g. https://www.cochranelibrary.com/cdsr/doi/10.1002/14651858.CD001431.pub5/full, https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9208236/) Methods -Under inclusion criteria, please expand what the study classify and accepts as SDM and how this was done in practice to allow reproducibility. -The study design criteria in PROSPERO seems to suggest that pre-post studies can be included too. Could you clarify if any change to the criteria has been made post registration and present it as part of the Methods section? -Please confirm the references for the methods section. Ref. 30 to the Cochrane Risk of Bias tool does not seem to link to the Cochrane Handbook. -Ref. 32 also does not relate to an explanation regarding the method of narrative synthesis used for reporting the results of this study.
-Was any hand searching done of the references of included studies?

Results
-Suggest noting the metrics used when reporting results to improve clarity. E.g. a significantly greater increase occurred in the intervention group alone [between group difference: 5.7 points; p<0.05] -> Which metric is used to ascertain this difference? How clinically meaningful is this?
Discussion and Conclusion -The discussion section seems to note that 2 studies (Gagne and Au) did not report a "a period of deliberation between patients and healthcare professionals, suggesting the decision-making process was not shared". Does this meant that the two studies did not include a shared decision making intervention? Please clarify this and the definition under Methods as well as confirm whether all studies fit the inclusion criteria.

Introduction
Could done in practice to allow reproducibility.
"For the purpose of this review, interventions were accepted if they were patient-facing and purposely stated that were to support SDM between a patient and healthcare professional." Methods The study design criteria in PROSPERO seems to suggest that pre-post studies can be included too. Could you clarify if any change to the criteria has been made post registration and present it as part of the Methods section?
An additional sentence has been added to the methods section to give examples of the types of studies included. This did include pre-post studies (see lines 190-191). "This included cluster trials, interrupted time series (ITS) studies with at least one data point before and after the intervention and controlled before-after (CBA) studies. " Methods Please confirm the references for the methods section. Ref. 30 to the Cochrane Risk of Bias tool does not seem to link to the Cochrane Handbook.
The Cochrane Handbook reference has been added here alongside the ROB-2 tool publication (see lines 220).

Methods
Ref. 32 also does not relate to an explanation regarding the method of narrative synthesis used for reporting the results of this study.
A further sentence has been added to this section to show the role of the reference in guiding the narrative synthesis (see lines 241-242). "Guidance on the conduct of narrative synthesis in systematic reviews [34] was consulted to guide this process." Methods Was any hand searching done of the references of included studies?
Yes. An additional sentence has been added to the Methods section (Information sources) to describe this (see lines [198][199]. "Reference lists of selected studies and relevant systematic reviews were also searched to identify further articles for inclusion."

Results
Suggest noting the metrics used when reporting results to improve clarity. E.g. a significantly greater increase occurred in the intervention group alone [between group difference: 5.7 points; p<0.05] -> Which metric is used to ascertain this difference? How clinically meaningful is this?
We have added metrics into the results section to make this clearer (see lines 280-375). To show the clinical significance of the results these are displayed in Table 1 alongside the properties of each measurement tool.

Discussion and conclusion
The discussion section seems to note that 2 studies (Gagne and Au) did not report a "a period of deliberation between patients and healthcare professionals, suggesting the decision-making process was not shared". Does this meant that the two studies did not include a shared decision making intervention? Please clarify this and the definition under Methods as well as confirm whether all studies fit the inclusion criteria.
Thank you for highlighting this needs to be clarified. The authors confirm that the interventions all met the eligibility criteria. A further sentence has been added to the Intervention section (see lines: 172-174). The appraisal in the discussion section is our interpretations of the interventions, using the MIND-IT model, and based upon the extracted data from the included studies manuscripts. "For the purpose of this review, interventions were accepted if they were patient-facing and purposely stated that were to support SDM between a patient and healthcare professional."

GENERAL COMMENTS
Author's Feedback An interesting study focusing on Shared Decision Making which needs to be implemented more in practice including in the CRD space. Some comments to improve the reporting of the study further: -Could you please expand the rationale on evaluating SDM interventions for CRD as a whole as under Introduction it was noted even for the same disease group e.g. asthma and cystic fibrosis, the evidence base found was too heterogeneous for meaningful conclusions to be reached? -While not the same, there may be benefits in looking into recent reviews on patient education materials which in many cases were part of a multicomponent intervention including face-to-face education, similar to the components defined as SDM in the study.

Introduction
Could you please expand the rationale on evaluating SDM interventions for CRD as a whole as under Introduction it was noted even for the same disease group e.g. asthma and cystic fibrosis, the evidence base found was too heterogeneous for meaningful conclusions to be reached?
There were very few studies included in the asthma review and zero studies in the cystic fibrosis review and therefore we have added this limited dataset to the text to show another rationale for exploring the efficacy of shared decision making across chronic respiratory disease.
Lines 145-150: Two systematic reviews have been published to synthesise evidence on the impact of SDM interventions on people with asthma[27] and cystic fibrosis[28] but they found the evidence base to be very limited and too heterogeneous to draw meaningful conclusions about the relationship between the interventions and improved SDM or downstream health-related outcomes and nearly all the review population were children and adolescents. Introduction While not the same, there may be benefits in looking into recent reviews on patient education materials which in many cases were part of a multicomponent intervention including face-to-face education, We have added an extra sentence to highlight this.

Results
Please include under Results reference to potentially relevant studies that were excluded as noted in the PRISMA checklist.
Additional details provided.
Lines 254-255: Other reasons included study design, type of intervention, and type of publication.

Results
The goal of a SDM tool is to support a patient to make a decision. Could you please expand for each included study what was the decision the tool was supposed to support? This can be part of Table 1. Were any for PR as it seems to be the topic of interest to the research group?
Thank you for this suggestion. We agree it would provide even more clarity and so have added a column into Table 1 (see  Line 652). Sadly we found no interventions to support PR decision making (see Lines 453-454).

Discussion
Under discussion, considering that most studies were feasibility/pilots and were not of high quality, would suggest clearly noting this caveat when mentioning how SDM improves quality of communication etc.
Have added this caveat to the beginning of the discussion.
Lines 384-386: Additionally, three studies measured intervention feasibility meaning they were unable to observe intervention efficacy [36,37,42], two RCTs were judged to have a high risk of bias [37,40] and two non-RCTs were judged to have a serious risk of bias [36,42].